Journal of Educational Psychology vol:84 issue:1 pages:85-94
According to Lewis and Mayer's (1987) simulation model for understanding compare problems, students make more comprehension errors when the order of the terms in the relational statement is not consistent with the preferred order. Three eye movement experiments, designed to test a number of hypotheses directly derived from this model, are discussed. The 1st experiment with university students solving 1-step compare problems revealed no evidence in favor of the model; the data of the 2nd experiment with 3rd graders, on the other hand, provided good support. To explain the results of the 1st experiment, a 3rd experiment was carried out in which university students were given a set of 2-step compare problems. The results from that study also fit with the model well, suggesting that the model holds true only when the task puts some cognitive demands on the subject.